Run with Docker

If you only want to run Next AI Draw.io on your own computer, Docker is the best way. This page shows how to start the app in a container, give it your AI provider settings, offer several models from the server and keep the admin panel settings.

Start the app

If Docker is not installed yet, install it first: Get Docker.

Then run:

docker run -d -p 3000:3000 \
  -e AI_PROVIDER=openai \
  -e AI_MODEL=gpt-4o \
  -e OPENAI_API_KEY=your_api_key \
  ghcr.io/dayuanjiang/next-ai-draw-io:latest

Open http://localhost:3000 in your browser.

The -e options set the AI provider, the model and the API key. Replace them with the settings of the AI provider you want to use. See AI providers for the options.

Use an env file

You can also keep the settings in an env file. Start from env.example, the template in the repository. If you cloned the repository, copy it to .env. Otherwise, download the file from GitHub:

# In a clone of the repository
cp env.example .env

# Without the repository
curl -o .env https://raw.githubusercontent.com/DayuanJiang/next-ai-draw-io/main/env.example

Edit .env with your settings, and start the container with --env-file:

docker run --env-file uses everything after the = as the value, including a comment at the end of the line. Some lines in env.example end with such a comment:

When you turn on a line like this by removing the # at its start, also delete the comment after the value:

The env_file of Docker Compose removes these comments by itself.

Offer several models from the server

You can mount an ai-models.json file into the container. With it, the server provides several models, and user API keys are not exposed. Mount the file at /app/ai-models.json:

docker run -d -p 3000:3000 \
  -e OPENAI_API_KEY=your_api_key \
  -v $(pwd)/ai-models.json:/app/ai-models.json:ro \
  ghcr.io/dayuanjiang/next-ai-draw-io:latest

To keep the file at a different path inside the container, set AI_MODELS_CONFIG_PATH to that path:

docker run -d -p 3000:3000 \
  -e OPENAI_API_KEY=your_api_key \
  -e AI_MODELS_CONFIG_PATH=/config/ai-models.json \
  -v $(pwd)/ai-models.json:/config/ai-models.json:ro \
  ghcr.io/dayuanjiang/next-ai-draw-io:latest

Keep the admin panel settings

The admin panel saves its settings in the /app/data folder inside the container. The docker-compose.yml in the repository mounts the ./data folder there (./data:/app/data), so the settings stay when the container is replaced. With docker run, add the same mount:

docker run -d -p 3000:3000 --env-file .env \
  -v $(pwd)/data:/app/data \
  ghcr.io/dayuanjiang/next-ai-draw-io:latest

The app in the container runs as the user nextjs (user ID 1001), and this user must be able to write to the folder. On Linux, if ./data does not exist, Docker creates it owned by root, and the admin panel becomes read only. Before you start the container, create the folder with the right owner. Run these commands where you run docker run or docker compose up:

mkdir -p data
sudo chown 1001:1001 data

If embed.diagrams.net is blocked

The app does not need embed.diagrams.net. The image on ghcr.io is built without NEXT_PUBLIC_DRAWIO_BASE_URL, and so is an image built from the docker-compose.yml in the repository. So the app serves its own copy of draw.io from /drawio, and the browser never contacts embed.diagrams.net. Only an image built with the build argument NEXT_PUBLIC_DRAWIO_BASE_URL=https://embed.diagrams.net loads draw.io from there.

To run the app without internet access, see Run it offline or on an intranet.